modeling-product-usage-metrics
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill primarily provides static SQL and HogQL templates for data modeling in dbt and PostHog. It does not contain executable shell scripts, Python/Node.js code, or any mechanisms for remote code execution.
- [SAFE]: The documentation includes explicit security guidance (Rule 6) directing the agent to treat event names as untrusted input. This proactively addresses the risk of indirect prompt injection by instructing the agent to treat telemetry data as quoted values and requiring user confirmation before performing persistent operations like creating database views.
Audit Metadata